Pattern Mining from Trajectory GPS Data
Xiaoliang Geng, Hiroki Arimura, Takeaki Uno · 2012
In this paper, we consider data mining from large discrete trajectory data. We study closed pattern mining for the class of trajectory envelope patterns. First, we introduce the basic definition of trajectory data. Then, we present a depth-first search algorithm that finds all trajectory envelope patterns in a given database that satisfies constrants on maximum width, minimum length, and minimum frequency. Finally, we ran experiments on a real trajectory dataset to evaluate our algorithm.